StarZero All shows Try the podcast clipper →

Mar 28, 2023 · 21m · mad

Not Just Another Cloud Database | SurrealDB Co-Founders Jaime & Tobie Morgan Hitchcock

Jaime & Tobie Morgan Hitchcock · 17m spoken Matt Turck · 9s spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

At Data Driven NYC, SurrealDB co-founders Jaime and Tobie Morgan Hitchcock present their unified multi-model database designed to simplify software architecture by combining document, graph, and relational capabilities with decoupled storage and direct client security.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 0.8% of the talking time here. How this is scored →

Matt as informed peer 0.0 Guest teaching 4.3 Guest disagreement 0.0 Matt pushing back 0.0
05100:0010:0020:000:15–2:36 · Matt as informed peer 0/10 Origins of SurrealDB and the Problem Statement Jaime and Tobie Morgan Hitchcock present their background building complex golf analytics software and explain the need for a unified multi-model database. The host does not participate as this is a presentation monologue.2:36–6:22 · Matt as informed peer 0/10 Architecture: Separating Storage and Compute Tobie presents the architecture of SurrealDB, highlighting storage-compute separation and detailing how multi-level graph queries operate without complex relational joins. The host remains silent during the presentation.6:22–10:55 · Matt as informed peer 0/10 Direct Client Access, Security, and Browser Integration Tobie details direct client access, embedded browser capabilities via IndexedDB, and granular row-level security permissions. The segment is entirely presenter-led with no host interaction.10:55–14:51 · Matt as informed peer 0/10 Future Roadmap: WebAssembly, Live Queries, Search & ML The presenters outline their technical roadmap for WebAssembly, live distributed queries, and machine learning vector embeddings on B-trees. The host does not speak during the monologue.0:15–2:36 · Guest teaching 3/10 Origins of SurrealDB and the Problem Statement Jaime and Tobie Morgan Hitchcock present their background building complex golf analytics software and explain the need for a unified multi-model database. The host does not participate as this is a presentation monologue.2:36–6:22 · Guest teaching 5/10 Architecture: Separating Storage and Compute Tobie presents the architecture of SurrealDB, highlighting storage-compute separation and detailing how multi-level graph queries operate without complex relational joins. The host remains silent during the presentation.6:22–10:55 · Guest teaching 4/10 Direct Client Access, Security, and Browser Integration Tobie details direct client access, embedded browser capabilities via IndexedDB, and granular row-level security permissions. The segment is entirely presenter-led with no host interaction.10:55–14:51 · Guest teaching 5/10 Future Roadmap: WebAssembly, Live Queries, Search & ML The presenters outline their technical roadmap for WebAssembly, live distributed queries, and machine learning vector embeddings on B-trees. The host does not speak during the monologue.0:15–2:36 · Guest disagreement 0/10 Origins of SurrealDB and the Problem Statement Jaime and Tobie Morgan Hitchcock present their background building complex golf analytics software and explain the need for a unified multi-model database. The host does not participate as this is a presentation monologue.2:36–6:22 · Guest disagreement 0/10 Architecture: Separating Storage and Compute Tobie presents the architecture of SurrealDB, highlighting storage-compute separation and detailing how multi-level graph queries operate without complex relational joins. The host remains silent during the presentation.6:22–10:55 · Guest disagreement 0/10 Direct Client Access, Security, and Browser Integration Tobie details direct client access, embedded browser capabilities via IndexedDB, and granular row-level security permissions. The segment is entirely presenter-led with no host interaction.10:55–14:51 · Guest disagreement 0/10 Future Roadmap: WebAssembly, Live Queries, Search & ML The presenters outline their technical roadmap for WebAssembly, live distributed queries, and machine learning vector embeddings on B-trees. The host does not speak during the monologue.0:15–2:36 · Matt pushing back 0/10 Origins of SurrealDB and the Problem Statement Jaime and Tobie Morgan Hitchcock present their background building complex golf analytics software and explain the need for a unified multi-model database. The host does not participate as this is a presentation monologue.2:36–6:22 · Matt pushing back 0/10 Architecture: Separating Storage and Compute Tobie presents the architecture of SurrealDB, highlighting storage-compute separation and detailing how multi-level graph queries operate without complex relational joins. The host remains silent during the presentation.6:22–10:55 · Matt pushing back 0/10 Direct Client Access, Security, and Browser Integration Tobie details direct client access, embedded browser capabilities via IndexedDB, and granular row-level security permissions. The segment is entirely presenter-led with no host interaction.10:55–14:51 · Matt pushing back 0/10 Future Roadmap: WebAssembly, Live Queries, Search & ML The presenters outline their technical roadmap for WebAssembly, live distributed queries, and machine learning vector embeddings on B-trees. The host does not speak during the monologue.

speaking balance: gold is Matt, purple is the guest (3 minute bins)

0:00 · Matt 0% · guest 100%0:00 · Matt 0% · guest 100%3:00 · Matt 0% · guest 100%3:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%6:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%9:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%12:00 · Matt 0% · guest 100%15:00 · Matt 4.2% · guest 95.8%15:00 · Matt 4.2% · guest 95.8%18:00 · Matt 1.4% · guest 98.6%18:00 · Matt 1.4% · guest 98.6%21:00 · Matt 57.6% · guest 42.4%21:00 · Matt 57.6% · guest 42.4%
Sharpest disagreement ▶ 18:30 Conceding essential database trade-offs

Tobie addresses the audience member's skeptical framing regarding performance trade-offs, acknowledging that SurrealDB had to drop standard features like auto-incrementing integer IDs.

 Hardest push from Matt ▶ 18:10 Audience question on performance trade-offs

An audience member challenges the pitch by pointing out that database design invariably forces trade-offs between performance, ease of use, and cost.

 Biggest teaching moment ▶ 12:35 Vector search mechanics on B-trees

Tobie educates the audience on how SurrealDB executes vector search without in-memory quad-trees by adapting embeddings to 2D B-trees due to storage-compute separation.

 Matt holds his own ▶ 15:36 Host taking control of the Q&A

Matt Turck asserts host authority at the end of the presentation, taking control of the floor to direct questions to new audience members.
the scores for every segment, with the reasoning behind each
ChapterTopicMatt as informed peerGuest teachingGuest disagreementMatt pushing backWhy
Origins of SurrealDB and the Problem Statement 0300 Jaime and Tobie Morgan Hitchcock present their background building complex golf analytics software and explain the need for a unified multi-model database. The host does not participate as this is a presentation monologue.
Architecture: Separating Storage and Compute 0500 Tobie presents the architecture of SurrealDB, highlighting storage-compute separation and detailing how multi-level graph queries operate without complex relational joins. The host remains silent during the presentation.
Direct Client Access, Security, and Browser Integration 0400 Tobie details direct client access, embedded browser capabilities via IndexedDB, and granular row-level security permissions. The segment is entirely presenter-led with no host interaction.
Future Roadmap: WebAssembly, Live Queries, Search & ML 0500 The presenters outline their technical roadmap for WebAssembly, live distributed queries, and machine learning vector embeddings on B-trees. The host does not speak during the monologue.

Statements from this episode (10)

Disclosure
SurrealDB founders relied on four databases for prior golf analytics platform
“In building this, we were building with InfluxDB, OrientDB, DynamoDB and MySQL.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 0:56
Insight
Graph data is the optimal way to query data going forward
“To be honest, I think graph data is really the way that we should be querying the data going forward. Ultimately, the world is not modeled according to tables or tabular data. It's modeled according to relationships and links between, between data.”
Tobie Morgan Hitchcock Mar 28, 2023 ▶ 4:21
Assertion Not checkable as stated
Multi-level graph queries are extremely hard to run in MongoDB or SQL
“So this type of query in a relational database or something like MongoDB is going to be very hard to run. You've got a lot of joins and you've got very complicated queries.”
Tobie Morgan Hitchcock Mar 28, 2023 ▶ 5:39
Assertion Supported
SurrealDB allows front-end applications to run without an API layer
“If people are building front-end applications, they can use SurrealDB completely without any middle layer or API layer sitting in front of it.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 6:32
Assertion Supported
SurrealDB can run embedded directly inside web browsers using IndexedDB
“Effectively, SurrealDB runs as an embedded database or in the cloud, but it enables us also to run directly in the browser in an embedded way on top of IndexDB.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 7:37
Assertion Not checkable as stated
Secondary streaming views allow real-time analytics without impacting database performance
“These allow streaming averages or walking averages and other calculations over your data in real time without affecting the performance of your database.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 9:35
Disclosure
SurrealDB will include WebAssembly plugins to enable custom logic and functions
“With WebAssembly plugins, which will be coming for our version one, that will enable all users to run their own custom logic and custom functions.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 11:19
Assertion Supported
Graph databases Neo4j and TypeDB lack concepts of tables or collections
“The graph database that people know, so NeoFj and TypeDB, they have no concept of tables. They have no concept of collections.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 17:22
Disclosure
SurrealDB decouples compute and storage using query nodes over key-value stores
“Our query nodes effectively are dumb nodes that sit in front of a key value store. That key value store can be embedded like RocksDB. It can be sitting in the browser, or it can be a distributed key value store that scales to terabytes or petabytes of data.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 18:50
Disclosure
SurrealDB rewrote its entire codebase from Go to Rust
“We decided to rewrite the whole thing in Rust because that gave us the guarantees A better language. Performance wasn't really the key, but enabling us to do other things with the language that we just couldn't do and go.”
Jaime & Tobie Morgan Hitchcock Mar 28, 2023 ▶ 20:24
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 conversations transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more